* HAL2 interrupt
@ 1998-10-27 23:59 Ulf Carlsson
1998-10-28 22:26 ` Thomas Bogendoerfer
0 siblings, 1 reply; 5+ messages in thread
From: Ulf Carlsson @ 1998-10-27 23:59 UTC (permalink / raw)
To: ralf; +Cc: linux
Hi,
I mailed you a few days ago, maybe that mail got dropped somewhere or maybe you
were to busy too read, anyhow ..
Now I'm asking again, do you know which interrupt number I should request for
the PBUS DMA:s. I can't handle the interrupts correctly without the knowledge of
interrupt numbers, and I'm afraid the spec didn't tell me anything about this
(atleast not the ones I've read). Also, is everything else I need for dealing
these interrupts ok (I haven't been able to check)? Maybe I could setup some
bogus irq detection stuff to figure this out on my own, but to save some bogus
programming I'm asking you instead.
It would be nice if you even explained *how* you know..
I think I'm pretty much done with a simple playback/recording driver now, still
untested though (probably +1000 serious bugs and race conditions, but I'll fix
that..).
Thanks,
Ulf
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: HAL2 interrupt
1998-10-27 23:59 HAL2 interrupt Ulf Carlsson
@ 1998-10-28 22:26 ` Thomas Bogendoerfer
1998-10-29 10:12 ` Ulf Carlsson
0 siblings, 1 reply; 5+ messages in thread
From: Thomas Bogendoerfer @ 1998-10-28 22:26 UTC (permalink / raw)
To: Ulf Carlsson, ralf; +Cc: linux
On Wed, Oct 28, 1998 at 12:59:01AM +0100, Ulf Carlsson wrote:
> Now I'm asking again, do you know which interrupt number I should request for
> the PBUS DMA:s. I can't handle the interrupts correctly without the knowledge of
for me it looks like SGINT_HPCDMA should be the first PBUS DMA interrupt.
But looking further into the indyIRQ.S and indy_int.c indicates, that these
interrupts aren't implemented, yet.
> It would be nice if you even explained *how* you know..
use the source Luke.
Thomas.
--
This device has completely bogus header. Compaq scores again :-|
It's a host bridge, but it should be called ghost bridge instead ;^)
[Martin `MJ' Mares on linux-kernel]
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: HAL2 interrupt
1998-10-28 22:26 ` Thomas Bogendoerfer
@ 1998-10-29 10:12 ` Ulf Carlsson
1998-10-29 22:10 ` Thomas Bogendoerfer
0 siblings, 1 reply; 5+ messages in thread
From: Ulf Carlsson @ 1998-10-29 10:12 UTC (permalink / raw)
To: Thomas Bogendoerfer, ralf; +Cc: linux
On Wed, Oct 28, 1998 at 11:26:52PM +0100, Thomas Bogendoerfer wrote:
> On Wed, Oct 28, 1998 at 12:59:01AM +0100, Ulf Carlsson wrote:
> > Now I'm asking again, do you know which interrupt number I should request for
> > the PBUS DMA:s. I can't handle the interrupts correctly without the knowledge of
>
> for me it looks like SGINT_HPCDMA should be the first PBUS DMA interrupt.
> But looking further into the indyIRQ.S and indy_int.c indicates, that these
> interrupts aren't implemented, yet.
The problem is that the sgiseeq and the sgiwd93 use HPC DMA as well, but they're
sharing local interrupt 0 (MIPS IRQ 2), and using some status register to
differentiate them. If I were writing a sgiwd93 driver or a sgiseeq driver, my
first guess would have been that they were using the SGINT_HPCDMA interrupts.
Well, looks like I'm out of luck, I'll do some trial & error.
Ralf?
- Ulf
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: HAL2 interrupt
1998-10-29 10:12 ` Ulf Carlsson
@ 1998-10-29 22:10 ` Thomas Bogendoerfer
1998-10-29 22:29 ` Ulf Carlsson
0 siblings, 1 reply; 5+ messages in thread
From: Thomas Bogendoerfer @ 1998-10-29 22:10 UTC (permalink / raw)
To: Ulf Carlsson, ralf; +Cc: linux
On Thu, Oct 29, 1998 at 11:12:11AM +0100, Ulf Carlsson wrote:
> The problem is that the sgiseeq and the sgiwd93 use HPC DMA as well, but they're
yes, but they aren't using PBUS DMA channels, because they aren't connected
to the PBUS.
> sharing local interrupt 0 (MIPS IRQ 2), and using some status register to
> differentiate them. If I were writing a sgiwd93 driver or a sgiseeq driver, my
> first guess would have been that they were using the SGINT_HPCDMA interrupts.
SCSI and ethernet are special HPC devices, so they get handled in a different
way.
> Well, looks like I'm out of luck, I'll do some trial & error.
try interrupt 12, which should be the HPC3 interrupt (sgint23.h). After
getting an HPC3 interrupt you have to look in HPC3 register istat0 and
istat1 for the interrupt source. I guess DaveM wanted to map these
interrupt sources to SGINT_HPCDMA, but never got to implement it.
Thomas.
--
This device has completely bogus header. Compaq scores again :-|
It's a host bridge, but it should be called ghost bridge instead ;^)
[Martin `MJ' Mares on linux-kernel]
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: HAL2 interrupt
1998-10-29 22:10 ` Thomas Bogendoerfer
@ 1998-10-29 22:29 ` Ulf Carlsson
0 siblings, 0 replies; 5+ messages in thread
From: Ulf Carlsson @ 1998-10-29 22:29 UTC (permalink / raw)
To: Thomas Bogendoerfer, ralf; +Cc: linux
> > Well, looks like I'm out of luck, I'll do some trial & error.
>
> try interrupt 12, which should be the HPC3 interrupt (sgint23.h). After
> getting an HPC3 interrupt you have to look in HPC3 register istat0 and
> istat1 for the interrupt source. I guess DaveM wanted to map these
> interrupt sources to SGINT_HPCDMA, but never got to implement it.
Thanks, I'll do this tonight or tomorrow.
By the way, I have patched the kernel to support ksymoops (with stack tracing,
call tracing and code dump), and Keith Owens has patched ksymoops to support
MIPS. There was some sort of show_registers() function in traps.c which wasn't
used. I rewrote that function and made it work (yes, the original function
didn't work at all). Sounds interesting, doesn't it? I'll send you the patch
tomorrow...
- Ulf
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~1998-10-29 22:28 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
1998-10-27 23:59 HAL2 interrupt Ulf Carlsson
1998-10-28 22:26 ` Thomas Bogendoerfer
1998-10-29 10:12 ` Ulf Carlsson
1998-10-29 22:10 ` Thomas Bogendoerfer
1998-10-29 22:29 ` Ulf Carlsson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox